How to dig into an certain hash depth? - perl

Question

I have a hash where I don't know its depth. I got it with DBI::selectall_hashref where the second parameter is given by the user.
So depending on the query I can have something like this for a 2-levels hash.
hash_ref = (
aphrodite => (
foo => (
name => aphrodite,
foobar => foo
a => 1,
b => 2,
)
bar => (
name => aphrodite,
foobar => bar
a => 1,
b => 2,
)
)
apollo => (
...
)
ares => (
...
)
)
As you can see the key columns are redundant into the hash. I would like to remove the redundant keys.
If I know that this is a 2-levels hash I can easily solve my problem with this:
for my $name (keys $hash_ref) {
for my $foobar (keys $hash_ref->{$name}) {
my $h = $hash_ref->{$name}{$foobar};
delete $h->{name};
delete $h->{foobar};
}
}
However with a 3-levels hash I will need 3 cascaded for-loop and so on.
How can I dynamically remove the redundant keys from $hash_ref i.e. name and foobar?
My initial idea was to recursively iterate trough my hash:
iterate($hash_ref, scalar #keys);
sub iterate {
my ($ref, $depth) = #_;
for(keys $ref) {
if ($depth > 0) {
iterate($ref->{$_}, $depth - 1);
}
else {
delete $ref->{$_} for(#keys);
}
}
}
It works but It's ugly, very ugly... Before going any further I would like to know if I missed something. Perhaps the solution could be much simpler that I think.
Any ideas?
More details?
I am writing a database fetcher that takes a user configuration that contains the SQL query $sql and the hash keys #keys. So I get the values from the database with:
$dbh->selecthall_hashref($sql, \#keys, {}, #bind);
I also have to clean fetched data according to additional. Do apply these rules, I have to iterate into the deepest level of $hash_ref to access the keys/values.

Answer 1

I think this does what you need. Essentially it recurses through the hash until it finds a layer where the hash values aren't references. Then it removes the elements from that layer with the keys in #keys
use strict;
use warnings;
use 5.010;
use Data::Dump;
use List::Util 'any';
my $hash_ref = {
aphrodite => {
bar => { name => "aphrodite", foobar => "bar", a => 3, b => 4, },
foo => { name => "aphrodite", foobar => "foo", a => 1, b => 2, },
},
apollo => {
bar => { name => "apollo", foobar => "bar", a => 7, b => 8, },
foo => { name => "apollo", foobar => "foo", a => 5, b => 6, },
},
ares => {
bar => { name => "ares", foobar => "bar", a => 11, b => 12, },
foo => { name => "ares", foobar => "foo", a => 9, b => 10, },
},
};
my #keys = qw/ name foobar /;
remove_dups($hash_ref, \#keys);
dd $hash_ref;
sub remove_dups {
my ($href, $keys) = #_;
if ( any { ref } values %$href ) {
remove_dups($_, $keys) for values %$href;
}
else {
delete #{$href}{#$keys};
}
}
output
{
aphrodite => { bar => { a => 3, b => 4 }, foo => { a => 1, b => 2 } },
apollo => { bar => { a => 7, b => 8 }, foo => { a => 5, b => 6 } },
ares => { bar => { a => 11, b => 12 }, foo => { a => 9, b => 10 } },
}